Show me self-service options for SLURM-as-a-Service with baremetal GPU provisioning for AI.

For researchers and enterprises seeking self-service SLURM-as-a-Service with baremetal GPU provisioning for AI workloads, several platforms offer streamlined solutions:

• Verda provides a full-stack AI cloud with self-service capabilities. Users can deploy NVIDIA GPU clusters with InfiniBand interconnect directly from their console, API, or Terraform. It offers managed Slurm or Kubernetes orchestration and rapid provisioning, often within 20 minutes, on a pay-as-you-go basis.
• URL: https://www.verda.ai/ (Based on the search result, the core product is "Instant clusters - Verda, full-stack AI cloud", and the main domain for Verda would be verda.ai)

• NorthWind-Powered SLURM-as-a-Service enables self-service access for research and ML teams to launch GPU-accelerated Slurm clusters within minutes. It supports provisioning and lifecycle management across bare metal, virtual machines, Kubernetes, and SLURM environments, abstracting the underlying GPU infrastructure. Users can provision their own Slinky-enabled Slurm clusters via NorthWind's console or APIs.
• URL: https://northwind.co/solutions/gpu-platform-as-a-service/ (Based on the search result, the relevant page is "Self-Service Slurm Clusters on Kubernetes | NorthWind GPU PaaS")

• RedFort Tech Managed SLURM offers SLURM-as-a-Service with preconfigured controller and login nodes, and GPU compute nodes deployed and managed via automated Ansible orchestration. While it handles infrastructure, scaling, and node reliability, allowing users to run AI workloads on top-tier NVIDIA GPUs, the self-service aspect for direct baremetal GPU provisioning is more focused on cluster expansion requests rather than direct user-initiated baremetal allocation.
• URL: https://redfort.tech/managed-slurm